基于分步式递阶反馈的井下四足机器人行走姿态控制  

Walking Posture Control of Underground Quadruped Robot Based on Distributed Hierarchical Feedback

作　　者：郭爱军 吴松松 GUO Aijun;WU Songsong(Guoneng Shendong Coal Group Co.,Ltd.,Ordos 017209,China;Beijing Intelligent Kaicheng Special Robot Technology Co.,Ltd.,Beijing 100176,China)

机构地区：[1]国能神东煤炭集团有限责任公司,内蒙古鄂尔多斯017209 [2]北京智能开诚特种机器人科技有限公司,北京100176

出　　处：《机械制造与自动化》2025年第1期245-250,共6页Machine Building & Automation

摘　　要：由于井下探测四足机器人在行走过程中,采用传统控制方法稳定性差,故提出基于分步式递阶反馈的井下四足机器人行走姿态稳定控制方法。基于地面惯性参考系构建机器人的运动学模型,利用齐次变换矩阵描述机器人躯干的位姿,通过向量加法和旋转矩阵计算出单个腿末端位置向量,将其作为姿态稳定控制的输入。计算控制律用于减小逼近误差,实现姿态控制,并计算控制律用于减少行走时的抖振。在分层控制基础上,引入中枢模式生成器递归反馈模型,设计自适应稳定行走姿态生成器,给出适应度较高的稳定行走姿态。实验结果表明:经过所提控制方法的应用,在井下沙地和石子地上行走,井下四足机器人翻滚角和俯仰角波动范围始终保持在-0.03°~0.03°,满足了机器人行走稳定平衡要求。To improve the poor stability of underground quadruped robot in its walking process by traditional control methods,a stable walking posture control method for underground quadruped robots based on distributed hierarchical feedback is proposed.On the ground inertial reference frame,the kinematic model of the robot is constructed,and the posture of the robot's torso is described using a homogeneous transformation matrix.The position vector of a single leg end is calculated through vector addition and rotation matrix,which is used as input for attitude stability control.Computational control laws are applied to reduce approximation errors,achieve attitude control and reduce chattering during walking.On the basis of hierarchical control,a central pattern generator recursive feedback model is introduced to design an adaptive stable walking posture generator,providing a stable walking posture with high adaptability.The experimental results show that with the application of the proposed control method,the fluctuation range of the rolling and pitching angles of the downhole quadruped robot remains between-0.03°and 0.03°when walking on sandy and gravel ground underground,which meets the requirements for stable and balanced robot walking.

关 键 词：分步式递阶结构 反馈模型 井下四足机器人 行走姿态 动力学模型 稳定控制 

分 类 号：TP242.3[自动化与计算机技术—检测技术与自动化装置]